Skip to content
ROOT
BRAND

Identity

The public-facing version of ROOT. The brand position, the audiences, the cadence behind it, what’s NOT promised. The doc the cinematic content + community channels point to.

If the Master Plan is the curriculum and The Story is the diary, this is the storefront. Abukix Studio is the name on the door — the public-facing layer that turns five years of pattern-first platform engineering into something a stranger can recognize, link to, and decide whether to follow. Everything in this doc is what the brand is, what it isn’t, and what shape it has to keep across all five years to stay coherent.

The brand is permanent; ROOT is one chapter under it. The headline doesn’t expire when the program ends — it rolls forward into whatever the next chapter is. That permanence is the whole point of writing the spine carefully now: the bios, the positioning, the audience definitions, the boundary rules. They have to survive a 5-year program plus whatever comes after, without rebrand.

The headline

Building an AI Platform in Public.

The brand spine. Permanent. Doesn’t expire. The work IS the brand: building, in public, an open AI platform that goes from kernel to LLM.

ROOT is the 5-year chapter that opens the brand — the arc from SRE to Staff/Principal AI Platform Engineer (Y1-Y5). Post-ROOT (Y6+), the brand continues under whatever role lands next (Staff/Principal, founder, applied AI lab, OSS maintainer). The headline rolls forward; ROOT was one chapter under it.

The bio

The brand spine, written in different lengths for different surfaces. Each is permanent — no dollar amounts, no timeframes baked in. A bio is brand spine, not progress ticker.

SurfaceBio
Permanent default
(GitHub profile, brand identity, conference bios)		Building an AI platform in public. Kernel to LLM.
LinkedIn headlineBuilding basecamp — an open AI platform, in public. Kernel to LLM.
YouTube AboutBuilding an AI platform in public. Documenting the work, the rhythm, and what it costs to do it honestly.
Instagram / X
(tagline-poster format)		An AI platform. Kernel to LLM. Built in public.

Optional: the “Year N” rolling format

For surfaces you can update annually (LinkedIn longer bio, Twitter/X bio):

Year N of building an AI platform in public. Kernel to LLM.

Updates take 30 seconds — increment the number each anniversary. Creates a returning-viewer hook (“oh, still doing it”). Post-ROOT it just becomes Year 6, 7, 8 — no rebrand. Use where bios are easy to update; skip on conference bios + printed surfaces.

Why this position works

The competitive frame. Most AI creators today focus on:

  • Prompt engineering
  • AI news / “10 AI tools you must use”
  • Wrapper-builders / no-code AI products

Far fewer document how to actually build and operate AI platforms. That’s where the SRE/platform background gives an edge: reliability, infrastructure, monitoring, automation, Kubernetes, operations — exactly the skills companies need for AI infrastructure. The lane is genuinely quieter than the prompt-engineering lane.

AWS Community Builder + CNCF Ambassador alignment. Content sits naturally in the overlap of both ecosystems:

AWS topicsCNCF topics
EKS, Bedrock, SageMaker, IAM, Terraform on AWSKubernetes, ArgoCD, Prometheus, OpenTelemetry, Kubeflow, KServe

You don’t choose one ecosystem. AI platforms sit right in the overlap. Both programs see content + OSS contributions + consistent cadence — exactly what ROOT (and post-ROOT) produces.

The mix: 80/15/5

Within the brand: 80% tech, 15% lifestyle, 5% faith.

The mix isn’t a content quota — it’s a check on tone. If a week’s worth of posts skews 50/50 tech and lifestyle, the brand has drifted toward influencer aesthetic. If faith content shows up more than 5% it stops being framing and starts being subject. The ratio is the guardrail, not the schedule.

80%: Core: Building the AI Platform

The journey through the work — content shaped around real artifacts:

  • Linux + Networking + Databases (Y1)
  • Kubernetes + GitOps + IaC (Y1-Y2)
  • AWS + GCP + Multi-cloud (Y2)
  • Platform Engineering + Service Mesh + Security (Y2)
  • Observability + eBPF + Lakehouse (Y3)
  • Stream + Batch + Data Serving + Governance (Y3)
  • MLOps + KServe + Ray + Feature Store + Kubeflow (Y4)
  • LLM Infrastructure + RAG + Vector DBs (Y4)
  • Agents + MCP + AIOps + Studio + mlship (Y5)
  • (Post-ROOT: whatever lands next)

Example titles:

  • “Day N of Building an AI Platform in Public”
  • “Deploying Kubeflow on Kubernetes”
  • “My MLOps Architecture on AWS”
  • “Building a RAG Pipeline from Scratch”
  • “What I Learned Migrating from SRE to AI Infrastructure”

15%: Lifestyle: the life behind the work

Visible but quiet:

  • Study sessions / homelab desk
  • Fitness progress (run, gym)
  • Bookshelf / reading list
  • Productivity systems
  • Behind-the-scenes content

This shows the rhythm (run-pray-build cadence) without making it the headline.

5%: Faith: occasional reflections

Sparse but present:

  • Discipline
  • Consistency
  • Service
  • Stewardship of time + craft
  • Purpose

Faith shows up as a frame (Sunday rhythm, “stewardship is paying attention”), not a topic on its own. Not preachy. Not proselytizing. Just present.

The three audiences

The 80/15/5 mix reaches three audiences via three discovery funnels.

1. Engineers (PRIMARY: 70%+ of content lands for them)

Who: junior to mid-career platform / SRE / data / ML / AI infrastructure engineers. How they find you: GitHub, Show HN, conference submissions, technical blog posts, LinkedIn, YouTube long-form. What they care about: patterns, depth, the specific architectural choices, the failure stories, the “I built this for 3 months — here’s what I’d do differently” content. What they stay for: basecamp, terralabs, mlship, platform-ctl as clone-and-self-host artifacts.

This is the audience AWS Community Builder / CNCF Ambassador programs serve.

2. Tech-shifters (SECONDARY)

Who: career-shifters from non-platform engineering backgrounds; aspiring engineers who want a model for “what does a deliberate platform-engineering life look like.” How they find you: Reels / YouTube Shorts / Instagram — the lifestyle-warm short-form. What they stay for: the inspiration + the roadmap. They may never run basecamp, but they internalize “this is what disciplined platform engineering looks like, day by day.”

3. Faith-first techies (TERTIARY)

Who: Christian engineers + faith-first techies looking for “craft as stewardship” framing. How they find you: word of mouth + the rare combination + Filipino Christian creator-economy networks. What they stay for: an integrated model — work, rhythm, faith aren’t separate compartments.

The brand spine: run-pray-build (the rhythm behind the 80%)

The 80% tech doesn’t run on willpower. It runs on a rhythm:

  • Run — body. Mornings and weekends. Outdoor when possible.
  • Pray — spirit. Sunday is the non-negotiable beat. Stewardship of time as Christian framing.
  • Build — craft. The 12 hrs/week of ROOT work, then post-ROOT whatever scale of work the next chapter calls for.

This triad is supporting infrastructure for the work, not the headline. The headline is the platform. The rhythm makes long-term commitment possible. Documented in The Story (The Rhythm section); appears in content as the 15% + 5% framing.

What Abukix Studio IS (technically)

A unified developer surface — Portal Web UI + command palette AI assistant + platform-ctl CLI + 5 documented composition recipes — layered on top of basecamp (the OSS platform). The visitor’s view of the work.

Public surface (Year 5 P29 launch): studio.abukix.dev.

Behind it:

  • 9-tier basecamp running on a homelab + 2 cloud clusters (substrate per homelab/hardware)
  • 10 OSS projects (basecamp, terralabs, platform-ctl, mlship, ops-handbook + Y1 fluency CLIs)
  • The pattern library (~50 patterns)
  • The 5-year operational journal

The technical product behind the brand is detailed in the Studio Plan, and the demos that anchor each launch live in Studio composition recipes.

What you ARE promising

  • The patterns (durable knowledge, ~50 entries by Y5)
  • The playbook (basecamp clone-and-bootstrap in <4 hours per Y5 success criteria)
  • The journey (5-year ROOT journal, weekly logs, postmortems, ADRs — the “build in public” surface)
  • The artifacts (10 OSS projects, ~30 upstream PRs across the program)
  • Consistent cadence — 1 YouTube long-form / week, 3-5 LinkedIn / week, daily Reels/Shorts (see content-playbook.md)

What you are NOT promising

  • Hosted compute at scale (the demo at studio.abukix.dev is rate-limited, CPU-only, 10-min sessions, $30-50/month cap — not a managed offering)
  • SLA / 24/7 support
  • “AI for everyone” end-user experience
  • Production-grade managed offering (deferred indefinitely)
  • Lifestyle aspiration without substance (lifestyle is 15%, not the headline)
  • Preachy faith content (faith is 5%, present in framing not subject)
  • Prompt-engineering / “10 AI tools” content (the saturated lane; not your brand)
  • A bio that expires (no dollar amounts, no timeframes baked into bio)

The 3-tier sharing model:

Tier 1: OSS, self-host             ← THE MOAT (free for everyone, free for you)
Tier 2: Hosted demo                ← THE BRAND MULTIPLIER ($30-50/mo cap)
Tier 3: Managed offering           ← DEFERRED INDEFINITELY (maybe never)

99% of users live in Tier 1. Tier 2 is the cinematic surface. Tier 3 is optional.

Boundary against IP risk

Day-job depth is real, but the public surface uses public sources only:

  • KubeCon talks, conference papers, OSS code, engineering blog posts published by companies
  • Not: internal designs, internal naming, internal architectural patterns from any current or past employer

Don’t reference any employer’s internal tooling in public content. The patterns are public; the implementations and the depth are public; specific internal company-isms are not. This includes the names of internal platforms, the shape of internal teams, and any architectural choices that aren’t documented in a public talk or paper.

The brand stands on its own — built from public sources, deployable on commodity hardware, transferable to any cloud. No “internal-vs-public” leakage.

Tone

Cinematic but specific. System as protagonist — voice-over narration, no talking head, hero shots of UI/terminal/system. Lifestyle B-roll (run, sunrise, Sunday log) wraps the tech but never replaces it. Faith framing is woven into Sunday rhythm + occasional stewardship reflection — never preachy.

References:

  • Apple product films for launch videos (system as protagonist, intentional pacing)
  • Tulio Sasaya production aesthetic for short-form (B-roll, floating UI overlays, subtitles) — minus the sponsor energy
  • Werner Herzog narration for year-end retrospectives (heard, not seen)
  • Cleo Abram for tech storytelling without creator-bro energy

Avoid:

  • Lifestyle aspiration without substance
  • Sponsor-driven tech-bro luxe aesthetic
  • Preachy / proselytizing faith framing
  • Prompt-engineering / “10 AI tools” content
  • Hot takes / motivational content / AI-generated polish

When the brand goes loud

Three loud-launch moments inside the ROOT chapter:

  1. Y2 P9 — terralabs ships publicly (first launch)
  2. Y3 P19 — basecamp goes public + blog post + LinkedIn
  3. Y5 P29 + P30 — Abukix Studio + platform-ctl + mlship v2 + pattern paper (the year the name lands)

Loud launches are produced (cinematic-tier video, multi-channel announce). Daily content stays low-production but consistent. Post-ROOT, additional loud-launch moments will emerge for whatever the next chapter ships — the brand is permanent, the launches keep coming.

Cross-references